Improving mortality prediction in Acute Pancreatitis by machine learning and data augmentation.
Computers in biology and medicine
Acute Pancreatitis (AP) is the inflammation of the pancreas that can be fatal or lead to further complications based on the severity of the attack. Early detection of AP disease can help save lives by providing utmost care, rigorous treatment, and better resources. In this era of data and technology, instead of relying on manual scoring systems, scientists are employing advanced machine learning and data mining models for the early detection of patients with high chances of mortality. The current work on AP mortality prediction is negligible, and the few studies that exist have many shortcomings and are impractical for clinical deployment. In this research work, we tried to overcome the existing issues. One main issue is the lack of high-quality public datasets for AP, which are crucial for effectively training ML models. The available datasets are small in size, have many missing values, and suffer from high class imbalance. We augmented three public datasets, MIMIC-III, MIMIC-IV, and eICU, to obtain a larger dataset, and experiments proved that augmented data trained classifiers better than original small datasets. Moreover, we employed emerging advanced techniques to handle underlying issues in data. The results showed that iterative imputer is best for filling missing values in AP data. It beats not only the basic techniques but also the Knn-based imputation. Class imbalance is first addressed using data downsampling; apparently, it gave decent results on small test sets. However, we conducted numerous experiments on large test sets to prove that downsampling in the case of AP produced misleading and poor results. Next, we applied various techniques to upsample data in two different class splits, a 50 to 50 and a 70 to 30 majority-minority class split. Four different tabular generative adversarial networks, CTGAN, TGAN, CopulaGAN, and CTAB, and a variational autoencoder, TVAE, were deployed for synthetic data generation. SMOTE was also utilized for data upsampling. The computational results showed that the Random Forest (RF) classifier outperformed all other classifiers on a 50 to 50 class split data generated by CTGAN, with 0.702 F and 0.833 recall. Results produced by RF on the TVAE dataset were also comparable, with 0.698 F. In the case of SMOTE-based upsampling, DNN performed best with a 0.671 F score.
A Nomogram With Six Variables Is Useful to Predict the Risk of Acquiring Carbapenem-Resistant Microorganism Infection in ICU Patients.
Frontiers in cellular and infection microbiology
Background:Carbapenem-resistant microorganism (CRO) transmission in the medical setting confers a global threat to public health. However, there is no established risk prediction model for infection due to CRO in ICU patients. This study aimed to develop a nomogram to predict the risk of acquiring CRO infection in patients with the first ICU admission and to determine the length of ICU stay (ICU-LOS) and 28-day survival. Methods:Patient data were retrieved from the Medical Information Mart for Intensive Care (MIMIC-IV) database based on predetermined inclusion and exclusion criteria. A CRO was defined as a bacterium isolated from any humoral microbial culture that showed insensitivity or resistance to carbapenems. The characteristics of CRO and non-CRO patients in the first ICU admission were compared. Propensity score matching was applied to balance the differences between the CRO and non-CRO cohorts. Kaplan-Meier curves were constructed to determine the 28-day survival rate and ICU-LOS. Furthermore, after randomization of the CRO cohort into the training and validation sets, a predictive nomogram was constructed based on LASSO regression and Logistic regression analysis, and its performance was verified by internal validation. Results:Overall, 4531 patients who had first ICU admission as recorded in MIMIC-IV were enrolled, 183 (4.04%) of whom were diagnosed with CRO infection. Moreover, CRO infection was independently associated with 28-day survival and ICU-LOS in ICU patients. Parameters eligible for inclusion in this nomogram were male sex, hemoglobin-min, temperature-max, use of a peripherally inserted central catheter line, dialysis treatment, and use of carbapenems. This nomogram showed a better performance as indicated by the area under the receiver operating characteristic curve values of 0.776 (95% confidence interval [CI] 0.667-0.750) and 0.723 (95% CI 0.556-0.855) in the training and validation sets, respectively, in terms of predicting the risk of acquiring CRO infection. Conclusions:CRO infection was independently associated with ICU-LOS and 28-day survival in patients with first ICU admission. The nomogram showed the best prediction of the risk of acquiring CRO infection in ICU patients. Based on the nomogram-based scoring, we can management the risk factors and guide individualized prevention and control of CRO.
Application of Machine Learning for Clinical Subphenotype Identification in Sepsis.
Infectious diseases and therapy
INTRODUCTION:Sepsis is a heterogeneous clinical syndrome. Identification of sepsis subphenotypes could lead to allowing more precise therapy. However, there is a lack of models to identify the subphenotypes in such patients. Thus, we aimed to identify possible subphenotypes and compare the clinical outcomes for subphenotypes in a large sepsis cohort. METHODS:This machine learning-based, cluster analysis was performed using the Medical Information Mart in Intensive Care (MIMIC)-IV database. We enrolled all adult (> 18 years old) patients diagnosed with sepsis in the first 24 h after intensive care unit (ICU) admission. K-means cluster analysis was performed to identify the number of classes. Multivariable logistic regression models were used to estimate the association between sepsis subphenotypes and in-hospital mortality. RESULTS:A total of 8817 participants with sepsis were enrolled. The median age was 66.8 (IQR, 55.9-77.1) years, and 38.1% (3361/8817) were female. Two subphenotypes resulted in optimal separation including 11 routinely available clinical variables obtained during the first 24 h after ICU admission. Participants in subphenotype B showed higher levels of lactate, glucose and creatinine, white blood cell count, sodium and heart rate and lower body temperature, platelet count, systolic blood pressure, hemoglobin and PaO/FiO ratio. In addition, the in-hospital mortality in patients with subphenotype B was significantly higher than that in subphenotype A (29.4% vs. 8.5%, P < 0.001). The difference was still significant after adjustment for potential covariates (adjusted OR 2.214; 95% CI 1.780-2.754, P < 0.001). CONCLUSIONS:Two sepsis subphenotypes with different clinical outcomes could be rapidly identified using the K-means clustering analysis based on routinely available clinical data. This finding may help clinicians to identify the subphenotype rapidly at the bedside.
Conservative oxygen therapy in critically ill and perioperative period of patients with sepsis-associated encephalopathy.
Frontiers in immunology
Objectives:Sepsis-associated encephalopathy (SAE) patients in the intensive care unit (ICU) and perioperative period are administrated supplemental oxygen. However, the correlation between oxygenation status with SAE and the target for oxygen therapy remains unclear. This study aimed to examine the relationship between oxygen therapy and SAE patients. Methods:Patients diagnosed with sepsis 3.0 in the intensive care unit (ICU) were enrolled. The data were collected from the Medical Information Mart for Intensive Care IV (MIMIC IV) database and the eICU Collaborative Research Database (eICU-CRD) database. The generalized additive models were adopted to estimate the oxygen therapy targets in SAE patients. The results were confirmed by multivariate Logistic, propensity score analysis, inversion probability-weighting, doubly robust model, and multivariate COX analyses. Survival was analyzed by the Kaplan-Meier method. Results:A total of 10055 patients from eICU-CRD and 1685 from MIMIC IV were included. The incidence of SAE patients was 58.43%. The range of PaO (97-339) mmHg, PaO/FiO (189-619), and SO≥93% may reduce the incidence of SAE, which were verified by multivariable Logistic regression, propensity score analysis, inversion probability-weighting, and doubly robust model estimation in MIMIC IV database and eICU database. The range of PaO/FiO (189-619) and SO≥93% may reduce the hospital mortality of SAE were verified by multivariable COX regression. Conclusions:SAE patients in ICU, including perioperative period, require conservative oxygen therapy. We should maintain SO≥93%, PaO (97-339) mmHg and PaO/FiO (189-619) in SAE patients.
Effects of Gastric Acid Secretion Inhibitors for Ventilator-Associated Pneumonia.
Frontiers in pharmacology
This study analyzed the association of gastric acid secretion inhibitors (GASIs) [including proton pump inhibitors (PPIs) and histamine 2 receptor antagonists (H2RAs)] with the occurrence of ventilator-associated pneumonia (VAP) and in-hospital mortality in patients who received invasive mechanical ventilation (IMV). Patients who received IMV and used GASI were included based on records in the MIMIC-IV database. The relationships of GASIs with VAP and the in-hospital mortality were determined using univariate and multivariate logistic regression analyses. Also, the effects of GASIs in some subgroups of the population were further analyzed. A total of 18,669 patients were enrolled, including 9191 patients on H2RAs only, 6921 patients on PPIs only, and 2557 were on a combination of the two drugs. Applying logistic regression to the univariate and multivariate models revealed that compared with H2RAs, PPIs had no significant effect on the incidence of VAP, and the combination of H2RAs and PPIs was a risk factor for VAP. Compared with H2RAs, univariate logistic regression revealed that, PPIs and combine the two drugs were both risk factors for in-hospital mortality, but multivariate logistic regression showed that they were not significantly associated with in-hospital mortality. In subgroup analysis, there were interaction in different subgroups of age, PCO2, myocardial infarct, congestive heart failure (P for interaction<0.05). Compared with H2RAs, PPIs did not have a significant association with either VAP or in-hospital mortality; the combination of H2RAs and PPIs was risk factor for VAP, but did not have a significantly associated with in-hospital mortality.
Clinical characteristics and risk factors associated with ICU-acquired infections in sepsis: A retrospective cohort study.
Frontiers in cellular and infection microbiology
Intensive care unit (ICU)-acquired infection is a common cause of poor prognosis of sepsis in the ICU. However, sepsis-associated ICU-acquired infections have not been fully characterized. The study aims to assess the risk factors and develop a model that predicts the risk of ICU-acquired infections in patients with sepsis. Methods:We retrieved data from the Medical Information Mart for Intensive Care (MIMIC) IV database. Patients were randomly divided into training and validation cohorts at a 7:3 ratio. A multivariable logistic regression model was used to identify independent risk factors that could predict ICU-acquired infection. We also assessed its discrimination and calibration abilities and compared them with classical score systems. Results:Of 16,808 included septic patients, 2,871 (17.1%) developed ICU-acquired infection. These patients with ICU-acquired infection had a 17.7% ICU mortality and 31.8% in-hospital mortality and showed a continued rise in mortality from 28 to 100 days after ICU admission. The classical Systemic Inflammatory Response Syndrome Score (SIRS), Sequential Organ Failure Assessment (SOFA), Oxford Acute Severity of Illness Score (OASIS), Simplified Acute Physiology Score II (SAPS II), Logistic Organ Dysfunction Score (LODS), Charlson Comorbidity Index (CCI), and Acute Physiology Score III (APS III) scores were associated with ICU-acquired infection, and cerebrovascular insufficiency, Gram-negative bacteria, surgical ICU, tracheostomy, central venous catheter, urinary catheter, mechanical ventilation, red blood cell (RBC) transfusion, LODS score and anticoagulant therapy were independent predictors of developing ICU-acquired infection in septic patients. The nomogram on the basis of these independent predictors showed good calibration and discrimination in both the derivation (AUROC = 0.737; 95% CI, 0.725-0.749) and validation (AUROC = 0.751; 95% CI, 0.734-0.769) populations and was superior to that of SIRS, SOFA, OASIS, SAPS II, LODS, CCI, and APS III models. Conclusions:ICU-acquired infections increase the likelihood of septic mortality. The individualized prognostic model on the basis of the nomogram could accurately predict ICU-acquired infection and optimize management or tailored therapy.
Clinical applications of machine learning in the survival prediction and classification of sepsis: coagulation and heparin usage matter.
Journal of translational medicine
BACKGROUND:Sepsis is a life-threatening syndrome eliciting highly heterogeneous host responses. Current prognostic evaluation methods used in clinical practice are characterized by an inadequate effectiveness in predicting sepsis mortality. Rapid identification of patients with high mortality risk is urgently needed. The phenotyping of patients will assistant invaluably in tailoring treatments. METHODS:Machine learning and deep learning technology are used to characterize the patients' phenotype and determine the sepsis severity. The database used in this study is MIMIC-III and MIMIC-IV ('Medical information Mart for intensive care') which is a large, public, and freely available database. The K-means clustering is used to classify the sepsis phenotype. Convolutional neural network (CNN) was used to predict the 28-day survival rate based on 35 blood test variables of the sepsis patients, whereas a double coefficient quadratic multivariate fitting function (DCQMFF) is utilized to predict the 28-day survival rate with only 11 features of sepsis patients. RESULTS:The patients were grouped into four clusters with a clear survival nomogram. The first cluster (C_1) was characterized by low white blood cell count, low neutrophil, and the highest lymphocyte proportion. C_2 obtained the lowest Sequential Organ Failure Assessment (SOFA) score and the highest survival rate. C_3 was characterized by significantly prolonged PTT, high SIC, and a higher proportion of patients using heparin than the patients in other clusters. The early mortality rate of patients in C_3 was high but with a better long-term survival rate than that in C_4. C_4 contained septic coagulation patients with the worst prognosis, characterized by slightly prolonged partial thromboplastin time (PTT), significantly prolonged prothrombin time (PT), and high septic coagulation disease score (SIC). The survival rate prediction accuracy of CNN and DCQMFF models reached 92% and 82%, respectively. The models were tested on an external dataset (MIMIC-IV) and achieved good performance. A DCQMFF-based application platform was established for fast prediction of the 28-day survival rate. CONCLUSION:CNN and DCQMFF accurately predicted the sepsis patients' survival, while K-means successfully identified the phenotype groups. The distinct phenotypes associated with survival, and significant features correlated with mortality were identified. The findings suggest that sepsis patients with abnormal coagulation had poor outcomes, abnormal coagulation increase mortality during sepsis. The anticoagulation effects of appropriate heparin sodium treatment may improve extensive micro thrombosis-caused organ failure.
Interpretable Machine Learning for Early Prediction of Prognosis in Sepsis: A Discovery and Validation Study.
Infectious diseases and therapy
INTRODUCTION:This study aimed to develop and validate an interpretable machine-learning model based on clinical features for early predicting in-hospital mortality in critically ill patients with sepsis. METHODS:We enrolled all patients with sepsis in the Medical Information Mart for Intensive Care IV (MIMIC-IV, v.1.0) database from 2008 to 2019. Lasso regression was used for feature selection. Seven machine-learning methods were applied to develop the models. The best model was selected based on its accuracy and area under curve (AUC) in the validation cohort. Furthermore, we employed the SHapley Additive exPlanations (SHAP) method to illustrate the effects of the features attributed to the model, and to analyze how the individual features affect the output of the model, and to visualize the Shapley value for a single individual. RESULTS:In total, 8,817 patients with sepsis were eligible for participation, the median age was 66.8 years (IQR, 55.9-77.1 years), and 3361 of 8817 participants (38.1%) were women. After selection, 25 of a total 57 clinical parameters collected on day 1 after ICU admission remained associated with prognosis and were used for developing the machine-learning models. Among seven constructed models, the eXtreme Gradient Boosting (XGBoost) model achieved the best performance with an AUC of 0.884 and an accuracy of 89.5% in the validation cohort. Feature importance analysis showed that Glasgow Coma Scale (GCS) score, blood urea nitrogen, respiratory rate, urine output, and age were the top 5 features of the XGBoost model with the greatest impact. Furthermore, SHAP force analysis illustrated how the constructed model visualized the individualized prediction of death. CONCLUSIONS:We have demonstrated the potential of machine-learning approaches for predicting outcome early in patients with sepsis. The SHAP method could improve the interpretability of machine-learning models and help clinicians better understand the reasoning behind the outcome.
Thiamine May Be Beneficial for Patients With Ventilator-Associated Pneumonia in the Intensive Care Unit: A Retrospective Study Based on the MIMIC-IV Database.
Frontiers in pharmacology
Ventilator-associated pneumonia (VAP) is a common infection complication in intensive care units (ICU). It not only prolongs mechanical ventilation and ICU and hospital stays, but also increases medical costs and increases the mortality risk of patients. Although many studies have found that thiamine supplementation in critically ill patients may improve prognoses, there is still no research or evidence that thiamine supplementation is beneficial for patients with VAP. The purpose of this study was to determine the association between thiamine and the prognoses of patients with VAP. This study retrospectively collected all patients with VAP in the ICU from the Medical Information Mart for Intensive Care-IV database. The outcomes were ICU and in-hospital mortality. Patients were divided into the no-thiamine and thiamine groups depending upon whether or not they had received supplementation. Associations between thiamine and the outcomes were tested using Kaplan-Meier (KM) survival curves and Cox proportional-hazards regression models. The statistical methods of propensity-score matching (PSM) and inverse probability weighting (IPW) based on the XGBoost model were also applied to ensure the robustness of our findings. The study finally included 1,654 patients with VAP, comprising 1,151 and 503 in the no-thiamine and thiamine groups, respectively. The KM survival curves indicated that the survival probability differed significantly between the two groups. After multivariate COX regression adjusted for confounding factors, the hazard ratio (95% confidence interval) values for ICU and in-hospital mortality in the thiamine group were 0.57 (0.37, 0.88) and 0.64 (0.45, 0.92), respectively. Moreover, the results of the PSM and IPW analyses were consistent with the original population. Thiamine supplementation may reduce ICU and in-hospital mortality in patients with VAP in the ICU. Thiamine is an inexpensive and safe drug, and so further clinical trials should be conducted to provide more-solid evidence on whether it improves the prognosis of patients with VAP.
The Association Between Bronchoscopy and the Prognoses of Patients With Ventilator-Associated Pneumonia in Intensive Care Units: A Retrospective Study Based on the MIMIC-IV Database.
Frontiers in pharmacology
In intensive care units (ICUs), the morbidity and mortality of ventilator-associated pneumonia (VAP) are relatively high, and this condition also increases medical expenses for mechanically ventilated patients, which will seriously affect the prognoses of critically ill patients. The purpose of this study was to determine the impact of bronchoscopy on the prognosis of patients with VAP undergoing invasive mechanical ventilation (IMV). This was a retrospective study based on patients with VAP from the Medical Information Mart for Intensive Care IV database. The outcomes were ICU and in-hospital mortality. Patients were divided based on whether or not they had undergone bronchoscopy during IMV. Kaplan-Meier (KM) survival curves and Cox proportional-hazards regression models were used to analyze the association between groups and outcomes. Propensity score matching (PSM) and propensity score based inverse probability of treatment weighting (IPTW) were used to further verify the stability of the results. The effect of bronchoscopy on prognosis was further analyzed by causal mediation analysis (CMA). This study enrolled 1,560 patients with VAP: 1,355 in the no-bronchoscopy group and 205 in the bronchoscopy group. The KM survival curve indicated a significant difference in survival probability between the two groups. The survival probabilities in both the ICU and hospital were significantly higher in the bronchoscopy group than in the no bronchoscopy group. After adjusting all covariates as confounding factors in the Cox model, the HRs (95% CI) for ICU and in-hospital mortality in the bronchoscopy group were 0.33 (0.20-0.55) and 0.40 (0.26-0.60), respectively, indicating that the risks of ICU and in-hospital mortality were 0.67 and 0.60 lower than in the no-bronchoscopy group. The same trend was obtained after using PSM and IPTW. CMA showed that delta-red blood cell distribution width (RDW) mediated 8 and 7% of the beneficial effects of bronchoscopy in ICU mortality and in-hospital mortality. Bronchoscopy during IMV was associated with reducing the risk of ICU and in-hospital mortality in patients with VAP in ICUs, and this beneficial effect was partially mediated by changes in RDW levels.
Outcomes of hyperlactatemia on admission in critically ill patients with acute myocardial infarction: A retrospective study from MIMIC-IV.
Frontiers in endocrinology
Background:It has not been verified whether there is a correlation between admission hyperlactatemia and outcomes in critically ill patients with acute myocardial infarction (AMI), especially in large data studies, which we aimed to do in this study. Methods:For this retrospective study, we extracted analysis data from a famous online intensive care unit database, the Medical Information Mart for Intensive Care (MIMIC)-IV. Included patients were divided into four groups according to the serum lactate level on admission. Hospital mortality and mortality over time were the main outcomes. To explore the relationship between admission hyperlactatemia and outcomes in critically ill patients with AMI, logistic regression, Cox regression, Kaplan-Meier curves, and subgroup analyses were used. Results:2171 patients matching the selection criteria were enrolled in this study. After adjusting for potential confounding factors, hyperlactatemia on admission contributed to increased short-term mortality in critically ill patients with AMI. The adjusted odds ratio for hospital mortality were 1.62, 3.46 and 5.28 in the mild, moderate, and severe hyperlactatemia groups (95% CI: 1.20-2.18, 2.15-5.58, and 2.20-12.70, respectively). The adjusted hazard ratio for 7-day and 30-day mortality were 1.99 and 1.35 (95% CI: 1.45-2.73 and 1.09-1.67) in the mild hyperlactatemia group, 3.33 and 2.31 (95% CI: 2.22-4.99 and 1.72-3.10) in the moderate hyperlactatemia group, 4.81 and 2.91 (95% CI: 2.86-8.08 and 1.88-4.50) in the severe hyperlactatemia group. The adjusted hazard ratio for 1-year and 5-year mortality were 2.03 and 1.93 (95% CI: 1.58-2.62 and 1.52-2.47) in the moderate hyperlactatemia group, 1.92 and 1.74 (95% CI: 1.28-2.89 and 1.17-2.59) in the severe hyperlactatemia group. Subgroup analyses indicated that the positive correlation between serum lactate level on admission and short-term mortality of critically ill patients with AMI was similar in the subgroups of cardiogenic shock and acute heart failure ( for interaction > 0.05). Conclusion:Hyperlactatemia, especially moderate and severe hyperlactatemia, on admission is closely related to higher short-term mortality incidence in critically ill patients with AMI. The relationship between serum lactate level on admission and short-term mortality of critical AMI patients is stable in subgroups of cardiogenic shock and acute heart failure.
Association between early ondansetron administration and in-hospital mortality in critically ill patients: analysis of the MIMIC-IV database.
Journal of translational medicine
BACKGROUND:While ondansetron (OND) is widespread availability, the contribution of OND to improve patient outcomes among intensive care unit (ICU) patients has not been examined. This study aimed to illustrate the association between early OND use and in-hospital mortality in critically ill patients and investigate whether this association differed according to OND dose. METHODS:The MIMIC-IV database was employed to identify patients who had and had not received OND. Statistical approaches included multivariate logistic regression, propensity score matching (PSM), and propensity score-based inverse probability of treatment weighting (IPTW) models to ensure the robustness of our findings. RESULTS:In total, 51,342 ICU patients were included. A significant benefit in terms of in-hospital mortality was observed in the OND patients compared to the non-OND group in the early stage [odds ratio (OR) = 0.75, 95% CI 0.63-0.89, p < 0.001]. In the circulatory system group, the early OND administration was associated with improved in-hospital mortality in ICU patients (OR 0.48, 95% CI 0.34-0.66; P < 0.001). The risk of in-hospital mortality was also lower in early OND users than in non-OND users both in the medical admission group and the surgical ICU admission group, and ORs were 0.57 (95% CI 0.42-0.76; P < 0.001) and 0.79 (95% CI 0.62-0.91; P < 0.001), respectively. A positive role of daily low- and moderate-dose OND treatment in early-stage was showed on the in-hospital mortality in PSM cohort, and the ORs were 0.75 (95% CI 0.62-0.90; P < 0.001) and 0.63 (95% CI 0.43-0.91; P < 0.001), respectively. The relationship between the daily low- and moderate-dose of OND and in-hospital mortality was also significant in ICU patients with cardiovascular diseases, and ORs were 0.51(95% CI 0.36-0.73; P < 0.001), and 0.26(95% CI 0.11-0.65; P < 0.001), respectively. Daily low-to-moderate dose of OND was also associated with in-hospital mortality in ICU entire cohort. CONCLUSIONS:Early OND use is closely associated with lower in-hospital mortality in ICU patients. Daily low-to-moderate dose of OND application is protective against in-hospital mortality. This association is more evident in the circulatory system group.
The Use of Antibiotics for Ventilator-Associated Pneumonia in the MIMIC-IV Database.
Frontiers in pharmacology
By analyzing the clinical characteristics, etiological characteristics and commonly used antibiotics of patients with ventilator-associated pneumonia (VAP) in intensive care units (ICUs) in the intensive care database. This study aims to provide guidance information for the clinical rational use of drugs for patients with VAP. Patients with VAP information were collected from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, including their sociodemographic characteristics, vital signs, laboratory measurements, complications, microbiology, and antibiotic use. After data processing, the characteristics of the medications used by patients with VAP in ICUs were described using statistical graphs and tables, and experiences were summarized and the reasons were analyzed. This study included 2,068 patients with VAP. Forty-eight patient characteristics, including demographic indicators, vital signs, biochemical indicators, scores, and comorbidities, were compared between the survival and death groups of VAP patients. Cephalosporins and vancomycin were the most commonly used. Among them, fourth-generation cephalosporin (ForGC) combined with vancomycin was used the most, by 540 patients. First-generati49n cephalosporin (FirGC) combined with vancomycin was associated with the highest survival rate (86.7%). More than 55% of patients were infected with Gram-negative bacteria. However, patients with VAP had fewer resistant strains (<25%). FirGC or ForGC combined with vancomycin had many inflammation-related features that differed significantly from those in patients who did not receive medication. Understanding antibiotic use, pathogenic bacteria compositions, and the drug resistance rates of patients with VAP can help prevent the occurrence of diseases, contain infections as soon as possible, and promote the recovery of patients.
Early combination of albumin with crystalloids administration might be beneficial for the survival of septic patients: a retrospective analysis from MIMIC-IV database.
Zhou Shiyu,Zeng Zhenhua,Wei Hongxia,Sha Tong,An Shengli
Annals of intensive care
BACKGROUND:Fluid therapy is a cornerstone in the treatment of sepsis. Recently, the guidelines have recommended the combined administration that using crystalloids plus albumin for septic patients, but the optimal timing for albumin combined is still unclear. The objective of this study was to investigate the association of timing of albumin combined with 28-day mortality in patients with sepsis. METHODS:We involved septic patients from the Medical Information Mart for Intensive Care (MIMIC)-IV database, and these patients were categorized into crystalloids group (crystalloids alone) and early combination group (crystalloids combined albumin at 0-24 h). The primary outcome was 28-day mortality. We used propensity score matching (PSM) to adjust confounding and restricted mean survival time (RMST) analysis was conducted to quantify the beneficial effect on survival due to the combination group. RESULTS:We categorized 6597 and 920 patients in the "crystalloids alone" and "early combination", respectively. After PSM, compared to the crystalloids group, the combination group was associated with the increased survival among 28-day (increased survival: 3.39 days, 95% CI 2.53-4.25; P < 0.001) after ICU admission. Patients who received albumin combination at the first 24-h was associated with prolonged LOS in ICU (10.72 days vs. 8.24 days; P < 0.001) but lower risk of 28-day mortality (12.5% vs 16.4%, P = 0.003) than those received crystalloids alone. CONCLUSION:In septic patients, receiving albumin combined within the first 24-h after crystalloids administration was associated with an increment of survival in 28 days.